Marker

Marker is an Erlang Markdown library, including parser and renderer.
The project has been created during the SpawnFest 2021
hackathon and consists of two parts:

• marker Erlang library in the marker/ directory exporting functions
  for parsing and rendering
• web demo in marker-web/ directory presenting the way Markdown documents
  are parsed (for e.g. debugging) and rendered into HTML.

How to run it?

As the marker is a library, it cannot be run on its own, but the build/test
instructions can be found under marker/ directory in its README.md file.

The demo project, marker_web is built in Elixir, uses Phoenix and can be run using
mix phx.server command. Detailed instructions, as above, can be found
in the marker_web/ directory.

How does it work?

The library, marker is build with a focus on the CommonMark specification
and uses hand written parser inspired by the CommonMark’s parsing strategy (because of time constraints it doesn’t follow it fully).

The website uses LiveView to pass input text to the parser and renderer,
gets resulting HTML and displays it immediately as a part of the page.

Parsing

The parsing process consists of two parts/steps:

1. During the first phase the input text is processed line-by-line and the overall
   structure of that is created, the tree of blocks (see specification for details).
2. During the second phase, the parser works on the leaves of the tree of blocks,
   so called inline elements - paragraphs and headings. In this step the parser
   interprets text formatting and links and after that the document tree is ready
   for rendering.

Rendering

Rendering process is much easier than parsing as it’s only generates strings
surrounded by proper HTML tags.

Renderer takes a document tree generated in the parsing phase, walks through that
and generated string representation of the children using the order stated
by parser.

Supported elements

Currenly marker (to some buggy extend) supports these Markdown elements:

• italics
• bolds
• headings
• blockquotes
• unordered lists
• horizontal rules
• inline code excerpts
• code blocks
• ordered lists

TBD:

• links
• images
• handling leading/trailing spaces
• handling UTF-8 strings
• nested formatting
• different edge cases
• more tests

Authors

The project has been created by the BKLH team of two:

• Burak @andreyuhai
• Łukasz @hryniuk